Fingerprint liveness detection consists in verifying if an input fingerprint image, acquired by a fingerprint verification system, belongs to a genuine user or is an artificial replica. Although several hardware- and software-based approaches have been proposed so far, this issue still remains unsolved due to the very high difficulty in finding effective features for detecting the fingerprint liveness. In this paper, we present a novel features set, based on the local phase quantization (LPQ) of fingerprint images. LPQ method is well-known for being insensitive to blurring effects, thus we believe it could be useful for detecting the differences between an alive and a fake fingerprint, due to the loss of information which may occur during the replica fabrication process. The method is tested on the four data sets of the Second International Fingerprint Liveness Detection Competition, and shows promising and competitive results with other state-of-the-art features sets. © 2012 ICPR Org Committee.

Fingerprint liveness detection consists in verifying if an input fingerprint image, acquired by a fingerprint verification system, belongs to a genuine user or is an artificial replica. Although several hardware- and software-based approaches have been proposed so far, this issue still remains unsolved due to the very high difficulty in finding effective features for detecting the fingerprint liveness. In this paper, we present a novel features set, based on the local phase quantization (LPQ) of fingerprint images. LPQ method is well-known for being insensitive to blurring effects, thus we believe it could be useful for detecting the differences between an alive and a fake fingerprint, due to the loss of information which may occur during the replica fabrication process. The method is tested on the four data sets of the Second International Fingerprint Liveness Detection Competition, and shows promising and competitive results with other state-of-the-art features sets.